Persistent microbial dysbiosis in preterm premature rupture of membranes from onset until delivery

Abstract

Background. Preterm Premature Rupture of Membranes (PPROM) is a major leading cause of preterm births. While the cause for PPROM remains unidentified, it is anticipated to be due to subclinical infection, since a large proportion of PPROM patients display signs of chorioamnionitis. Since subclinical infections can be facilitated by dysbiosis, our goal was to characterize the vaginal microbiome and amniotic fluid discharge upon PPROM, through latency antibiotic treatment, and until delivery, to detect the presence of pathogens, microbiota alteration, and microbial response to treatment. Methods. Enrolled subjects (15) underwent routine institutional antenatal care for PPROM, including the administration of latency antibiotics. Serial vaginal swabs were obtained from diagnosis of PPROM through delivery and the sequencing of the V3-V5 region of the 16S rRNA gene was performed for all collected samples. Results. The results show that Lactobacilli species were markedly decreased when compared to vaginal swabs collected from uncomplicated pregnancy subjects with a matched gestational time. Prevotella and Peptoniphilus were the most prevalent taxa in PPROM subjects at presentation. The vaginal microbiome of the PPROM subjects varied substantially intra- and inter-subjects. Several taxa were found to be significantly reduced during and after the antibiotic treatment: Weeksella, Lachnospira, Achromobacter, and Pediococcus. In contrast, Peptostreptococcus and Tissierellaceae ph2 displayed a significant increase after the antibiotic treatment. However, the relative abundance of Lactobacillus, Prevotella, and Peptoniphilus was not substantially impacted during the hospitalization of the PPROM subjects. The deficiency of Lactobacillus, and constancy of known pathogenic species, such as Prevotella and Peptoniphilus during and after antibiotics, highlights the persistent dysbiosis and warrants further investigation into mitigating approaches. Discussion. PPROM is responsible for one third of all preterm births. It is thought that subclinical infection is a crucial factor in the pathophysiology of PPROM because 25-40% of patients present signs of chorioamnionitis on amniocentesis. Here we sought to directly assess the bacterial content of the vagina and leaking amniotic fluid of subjects at presentation, throughout treatment and up until delivery, in order to search for common pathogens, microbiota changes, and microbial response to latency antibiotic treatment. We have found that the vaginal microbiome of PPROM subjects is highly variable and displays significant changes to treatment. However, the unchanging deficiency of Lactobacillus, and persistence of known pathogenic species, such as Prevotella and Peptoniphilus from presentation, through antibiotic treatment and up until delivery, highlights the persistent dysbiosis and warrants further investigation into mitigating approaches.

Keywords: Lactobacillus; Microbiome; Obstetrics; PPROM; Peptoniphilus; Prevotella.

Figure 1. Samples collected and analyzed in the course of the study.

Figure 2. Comparison of relative abundance of taxa at the genus level between vaginal swabs of subjects at the time of presentation at the emergency room with PPROM before the administration of antibiotic treatment (9 subjects) and vaginal swabs from subjects that underwent an uncomplicated pregnancy at approximately 29 weeks of gestation (12 subjects).

The lack of lactobacilli dominance in PPROM subjects is apparent, as is the inter-individual variation. Only taxa at >1% relative abundance in the PPROM subjects are shown for graphical clarity.

Figure 3. Boxplot representing the Lactobacillus frequency in the vaginal swabs of 9 PPROM subjects at presentation before the administration of antibiotic treatment.

Normal is represented by the vaginal swabs of 12 subjects with uncomplicated pregnancies at approximately 29 weeks of gestation. Statistical significance found through linear regression with permutation.

Figure 4. Summary (Median >1% relative abundance) of taxa at the genus level from all PPROM subjects at presentation before the administration of antibiotic treatment (9 subjects).

Figure 5. Placental and maternal microbiome taxa variation.

S, Subject; LD, Latency day; P, Placental sample.